Heuristic algorithms for the multi-period fixed charge models

Abstract

The multi-period fixed charge problem is an extension of the multiperiod distribution problem and general fixed charge transportation problem, where the time based decisions on the size of the shipments, simultaneous consideration of both suppliers and customers. Inventories and backorders/subcontracts can make an economical distribution. The conventional transportation problem considers only per unit cost of transportation. The other that has wide acceptance is FCT problem. Concerning the above, this thesis addresses four multiperiod fixed charge models. They are (1) Multi-period fixed charge distribution problem associated with backorder and inventories.(2) Multi-period fixed charge distribution problem associated with subcontract and inventories. (3) Multi-period fixed charge production-distribution problem associated with backorder and inventories. (4) Multi-period fixed charge production-distribution problem associated with subcontract and inventories. In recent years, problem specific simple heuristic algorithm, neighbourhood search based Simulated Annealing Algorithmand population search based Genetic Algorithm and have been increasingly applied to various search and optimization problems and have emerged as potential techniques to provide solutions with acceptable accuracy for NPhard problems. In the light of the above consideration, this thesis proposes Equivalent Variable Cost heuristic from simple problem specific heuristics, from neighbourhood search based heuristics and GA from population search based heuristics to solve the above four multi-period fixed charge models to minimize the total cost. The proposed heuristics are evaluated for their solution quality by comparing them with lower bound value and LINGO solutions. The comparison reveals that the proposed SAA and GA generate better solutions than the EVC heuristic solutions and are capable of providing solution either equal or close to the lower bound value and optimal solution of the problems.